DESCRIPTION (adapted from the Abstract): Several critical issues regarding the pathogenesis of HIV remain unanswered. In the proposed work the Investigator and her associates will explore the processes whereby HIV invades the CNS, establishes residency, and ultimately elicits neurological disease. Two distinct but related issues will be studied: (1) virus entry into the CNS; and (2) the contribution of HIV genotype to the development of HIV dementia. This research plan entails the use of the MACS, a cohort of homosexual men recruited in 1984, which includes seroconverters with known dates of infection, who are followed prospectively with detailed neuropsychological assessments to identify HIV. Banked plasma and CSF are available from different time points following seroconversion through death, and in many, autopsy tissue has been harvested using a systematic protocol. Molecular clones containing a 3.6-3.9 kb portion of the viral genome encoding the gp160 and nef genes and the 3' LTR will be generated and sequenced. Sequences from plasma and CSF specimens obtained early and late in infection and those subsequently found in postmortem brain will be compared. The Investigator hypothesizes that the disparity observed between HIV-1 genotypes in brain and blood results from the early seeding of the brain from trafficking cells rather than direct parenchymal infection from cell-free virions. Further, she anticipates that significant genotypic differences will be observed between demented and non-demented patients' brains, and that divergence will be observed among different regions of the same brain. Confirmation of early brain infection may have important implications for the timing of initiation of antiretroviral therapy in asymptomatic individuals.